Re: Bug in autovacuum.c?

On Apr 1, 2011, at 4:48 PM, Bruce Momjian wrote:
> I am not so concerned about this case but about other cases where we are
> computing xid distances across the invalid range.

Bruce, I think you hit the nail on the head earlier:

> To do the right thing every computation that passes over the xid
> wraparound bounary should subtract FirstNormalTransactionId, not just
> those that fall in the boundry.

Put another way: XID calculations should not use just +/-, but an operator (presumably a macro) that understands wraparound and the special values. Surely we have a similar problem in the code that increments XIDs, and possibly other places as well.
--
Jim C. Nasby, Database Architect                   jim@nasby.net
512.569.9461 (cell)                         http://jim.nasby.net